[gmx-users] MSD again

[Lars Poulsen]

Hi!
Once again I have to bother you all with more questions about g_msd and g_analyze, since I still don't think I understand the output of the programs.

As an illustration, I have attached an analysis of the same trajectory by both g_msd and g_analyze (jpg-files). It's a trajectory of one oxygen molecule in a glassy polymer.

The output of g_msd doesn't look like a MEAN SQUARE displacement (<|r(t)-r(0)|^2>) to me, it looks more like the simple displacement of the molecule (|r(t)-r(0)|). Or am I wrong?

g_analyze gives a nice smooth curve, but it only uses half the trajectory. When calculating the ACF, I can use the -acflen flag to change this, but it doesn't seem to work for the msd output. What am I doing wrong?

More important is the question: which program can I trust, g_analyze or g_msd? What are people's experiences with calculating MSD using either g_msd or g_analyze?

thanks,
mvh Lars 
-----------------------------------------------------------------
Lars Poulsen
Dept. of Chemistry, University of Aarhus
Langelandsgade 140, DK-8000 Aarhus C
Phone: +45 8942 3871
Fax: +45 8619 6199
email: lpoulsen at chem.au.dk
-----------------------------------------------------------------

-------------- next part --------------
A non-text attachment was scrubbed...
Name: g_msd.jpg
Type: image/jpeg
Size: 35598 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20020212/ce1c379f/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: g_analyze.jpg
Type: image/jpeg
Size: 27580 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20020212/ce1c379f/attachment-0001.jpg>